An Environmentally Friendly Method for As, Cd, Cr, Cu, Ni, and Pb Determination in Terrestrial Moss Samples Using Ultrasonic Slurry Sampling Combined with Electrothermal Atomic Absorption Spectrometry

1 An Environmentally Friendly Method for As, Cd, Cr, Cu, Ni, and 2 Pb Determination in Terrestrial Moss Samples Using Ultrasonic 3 Slurry Sampling Combined with Electrothermal Atomic 4 Absorption Spectrometry 5 6 M.C. Barciela Alonsoa, M. Noya Pardalb, M. López Pedrousob, J.R. Aboalb, 7 J.A. Fernándezb, and P. Bermejo-Barreraa 8 aDepartment of Analytical Chemistry, Nutrition and Bromatology, Faculty of 9 Chemistry, University of Santiago de Compostela, 15782, Santiago de 10 Compostela, Spain 11 b Ecology, Department of Cellular Biology and Ecology, Faculty of Biology, 12 University of Santiago de Compostela, 15782, Santiago de Compostela, Spain 13 14 *corresponding author: pilar[email protected] 15 16 17 Abstract: 18 Novel ultrasonic slurry sampling methods combined with the ETAAS 19 determination of As, Cd, Cr, Cu, Ni, and Pb in terrestrial moss samples were 20 developed without using corrosive acids. A pulverized moss sample between 1 to 21 50 mg (depending on element concentration) was placed into a 2.5 mL 22 autosampler cup with 2 mL of HNO3 (0.2 M), containing 0.01% of Triton X-100 23 reagent. The sample was then sonicated directly into the autosampler cup, using 24 an ultrasonic probe at 85% power setting of 7 W for 25 seconds before injection. 25 Chemical modifiers were introduced directly into the graphite tube after sample 26 introduction. An amount of 3 mu g of Mg(NO3)(2) and 5 mu g of Pd was used as 27 the chemical modifier for the As and Cu determination; 70 mu g of NH4H2PO4 28 and 4 mu g of Mg(NO3)(2) for Cd and Pb determination; and 20 mu g of 29 Mg(NO3)(2) for Cr. Nickel (Ni) was determined without the addition of a chemical 30 modifier. The methods were sensitive and precise, resulting in LODs of 51.3, 4.2, 31 94.4, 57.0, 124.7, 50.5 ng g(-1) for As, Cd, Cr, Cu, Ni and Pb, respectively, and 32 RSDs of <10%. The accuracy was evaluated using certified reference materials 33 GBW 07604 Poplar Leaves and Pseudoscleropodium purum (M2 and M3), and 34 the results obtained were in good agreement with the certified values for all the 35 elements studied. The developed method was applied for element determination 36 in real moss samples and to the best of our knowledge, no studies have 37 previously reported trace element determination in moss samples using this 38 methodology. 39 40 Keywords: 41 42 43 1. Introduction 44 Several plants (Lichens, tree leaves, shrubs, Tillandsia, and mosses) were used in 45 different studies for monitoring the atmospheric pollution (1–3). Among these, 46 terrestrial mosses are the most commonly used biomonitors for trace metal 47 atmospheric levels (4) and hundreds of scientific articles have been published (5–13). 48 Commission for Europe, New York, USA) (15). Every five years, sampling surveys 49 are performed across Europe, up to 4500 sites from 25 European countries in the last 50 survey (15). The high number of samples and the number of elements usually 51 determined in these studies requires reliable analytical techniques. The trace elements 52 included in ICP Vegetation reports are Al, As, Cd, Cr, Cu, Fe, Hg, Ni, Pb, V, and Zn 53 (14). In the present study, we selected As, Cd, Cr, Cu, Ni, and Pb, most of which are 54 included in the European Air Quality Directives. Several analytical techniques have 55 been used for the determination of these elements in moss samples, such as 56 electrothermal atomic absorption spectrometry (ETAAS), flame atomic absorption 57 spectrometry (FAAS), inductively coupled plasma atomic emission spectrometry 58 (ICP-AES), inductively coupled plasma mass spectrometry (ICPMS), instrumental 59 neutron activation analysis (INAA), and atomic fluorescence spectrometry (AFS) 60 (14). The use of these techniques varies among the studied elements. The most 61 commonly used analytical techniques in moss analysis are ICPMS and ICP-AES due 62 to their high sensitivity and multi-elemental analysis capacity. However, this type of 63 instrumentation is often too costly for some laboratories, while ETAAS is a technique 64 available in most analytical laboratories and also offers high sensitivity for moss 65 analysis, especially for As, Cd, and Pb determination (7,14,16–18). Microwave acid 66 sample digestion was usually performed prior to trace metal determination in moss 67 samples by ETAAS, ICP-OES, and ICP-MS (7–9,17). Nitric acid or mixtures of 68 HNO3 with H2O2 (7–9) were used for microwave acid digestion of moss samples for 69 trace metal determination. The main drawbacks of sample acid digestion are the use 70 of dangerous corrosive chemical products and the long time required for sample pre- 71 treatment. An alternative methodology to conventional wet digestion procedures is 72 the slurry sampling technique. This is a well-established methodology for the direct 73 determination of trace metals in solid samples by atomic spectrometry (19), providing 74 good analytical results and avoiding the use of dangerous and corrosive reagents. 75 Although this methodology has been widely used for trace metal determination in 76 several types of samples such as seafood (20,21), human hair (22), sediments and 77 sludges (21,23,24), ambient aerosols (25), and plant materials (3,21,26,27), to the best 78 of our knowledge, no studies have reported trace metal determination in moss samples 79 using slurry sampling combined with ETAAS. The objective of this study was to 80 develop sensitive, reliable, relatively rapid, and environmentally friendly methods for 81 As, Cd, Cr, Cu, Ni, and Pb determination in real terrestrial moss samples using 82 ultrasonic slurry sampling combined with ETAAS. 83 EXPERIMENTAL 84 Instrumentation 85 For this study, a PerkinElmer AAnalyst™ 600 atomic absorption spectrometer with 86 Zeeman background correction was used, equipped with an AS 800 autosampler, 87 provided with a USS 800 titanium ultrasonic probe (PerkinElmer, Inc., Shelton, CT, 88 USA). The radiation sources were hollow cathode lamps for Cd, Cu, Cr, Ni, and Pb 89 and an electrodeless discharge lamp operated from an external power supply for As. 90 The instrumental operating conditions are listed in Tables I and II. Pyrolytically 91 coated graphite tubes with L’vov platform were used for the elements studied. 92 Integrated absorbance with an integration time of 3 seconds was used for As, Cd, and 93 Pb, and 5 seconds for Cu, Cr, and Ni. The injected sample volume was 20 μL. An 94 ultracentrifuge mill ZM 200 (Retsch, Haan, Germany), equipped with titanium 95 components, was used to reduce the particle size of the moss samples (less than 50 96 μm). A microbalance Excelence Plus, XP 26 (± 0.001 mg) from Mettler Toledo® 97 (Greifensee, Switzerland) was used to weigh the samples. 98 Standard Solutions and Reagents 99 Stock standard solutions for As, Cd, Cu, Cr, Ni, and Pb of 1000 μg L-1 each were 100 obtained from Panreac (Spain). Each test solution was prepared with ultrapure water 101 (obtained from a Milli-Q® water-purification system, Millipore Corporation, USA) 102 immediately before use. Triton® X-100, polyethylene glycol mono(p-1,1,3,3,- 103 tetramethylphenyl) ether, was obtained from Merck (Darmstadt, Germany). 104 Palladium nitrate 10 g L-1 in HNO3 (15%) (Merck, Darmstadt, Germany), 105 magnesium nitrate (Merck), and NH4H2PO4 (99.999%) (Aldrich Chemical, 106 Milwaukee, WI, USA) were used as the chemical modifiers. Certified reference 107 material GBW 07604 Poplar Leaves (Institute of Geophysical and Geochemical 108 Exploration, Langfang, P.R. China) was used in the accuracy study. Also used were 109 two moss reference materials (Pleurozium scherberi) M2 and M3 (collected in 110 Harjavalta (61º22´N 22º12´E) in the vicinity of a metal smelter) which were obtained 111 in interlaboratory comparison studies with the University of Santiago de Compostela, 112 Spain (28). 113 Sample Preparation 114 Terrestrial moss samples were collected around a ferrous smelter (Meg), a non-ferrous 115 smelter (Alc), and in a background area (OSE). Collection of the moss samples was 116 performed according to the methods developed for this moss species (29–32). The 117 samples were picked by hand using gloves (talcum-free) and placed into polyethylene 118 bags. In the laboratory, the samples were stored at 4 ºC until sample processing (within 119 24-48 hours). After this, the apices of the shoots (3-4 cm) were cut from the moss 120 Pseudoscleropodium purum. The samples were then dried to constant weight in a Dry- 121 Big forced-air oven (Selecta) at 40 ºC. The dried samples were then pulverized in a 122 Retsch ZM200 metal-free mill (particle size of less than 100 μm, Retsch Haan, Germany) 123 and stored in a polyethylene vial (acid-washed with 10% v/v nitric acid for 24 hours) until 124 ETAAS analysis. 125 Procedure for Slurry Preparation 126 A mass of pulverized moss sample (between 1 and 50 mg) was placed into a 2.5 mL 127 autosampler cup containing 2 mL of HNO3 (0.2 M) and 0.01% of Triton X-100. The 128 sample was then sonicated using an ultrasonic probe (at 85% power setting – 7W) for 129 25 seconds before injection directly into the autosampler cup. 130 Element Determination by ETAAS 131 A volume of 20 μL of the slurry sample, prepared as described above, with 5 μL of the 132 respective chemical modifiers used for the different elements, were introduced into the 133 graphite furnace tube for analysis. Amounts of 3 μg Mg(NO3)2 and 5 μg Pd were the 134 chemical modifiers for As and Cu determination; 70 μg NH4H2PO4 and 4 μg Mg(NO3)2 135 for Cd and Pb determination; and 20 μg Mg(NO3)2 for Cr. Nickel was determined without 136 the addition of a chemical modifier. The instrumental conditions as well as the graphite 137 furnace programs used for the determination of these elements are listed in Tables I and 138 II. 139 RESULTS AND DISCUSSION 140 Optimization of Graphite Furnace Program 141 Experiments were performed to determine the optimum temperatures and times for the 142 drying, mineralization, and atomization steps. The study was performed using an 143 aqueous standard of the element studied and a moss slurry sample (prepared as 144 described in the Procedure section above) and using Pd-Mg(NO3)2 as the chemical 145 modifier for As and Cu, NH4H2PO4-Mg(NO3)2 for Cd and Pb, and Mg(NO3)2 for Cr. Two 146 drying steps (110 ºC and 130 ºC) were used (Table II). Determination of the optimum 147 mineralization temperatures were performed varying the mineralization temperature 148 from 900 to 1400 ºC for As, 500–1000 ºC for Cd, 1000–1700 ºC for Cr, 800–1500 ºC for 149 Cu, 800–1400 ºC for Ni, and 600–1100 ºC for Pb. Taking into account the results 150 obtained, 1300 ºC, 700 ºC, 1200 ºC, 1300 ºC, 1200 ºC, and 900 ºC were selected as the 151 optimum mineralization temperatures for As, Cd, Cu, Cr, Ni, and Pb, respectively (Table 152 III). 153 Optimization of the atomization temperatures was performed in the same way, and 154 working in the temperature ranges of 2000–2400 ºC for As, 1200–1600 ºC for Cd, 1600– 155 2300 ºC for Cu, 1500–2400 ºC for Cr, 1800–2400 ºC for Ni, and from 1300–1800 ºC for 156 Pb. The selected optimized temperatures were 2100 ºC, 1300 ºC, 1900 ºC, 2200 ºC, 157 2300 ºC, and 1500 ºC for As, Cd, Cu, Cr, Ni and Pb, respectively (Table III). These 158 temperatures were selected taking into account the integrated absorbance as well as 159 the peak profile. The influence of ramp times and hold times for each step was also 160 studied, and the optimized graphite furnace programs are listed in Table II. 161 Amount of Chemical Modifier 162 Selection of the chemical modifiers used for ETAAS determination of the elements was 163 based on previous studies (20,22,24,27,33–35) as well as per the manufacturer’s 164 (PerkinElmer, Inc.) recommendations (36,37). As listed above, Pd-Mg(NO3)2 was 165 selected as the chemical modifier for As and Cu determination, NH4H2PO4-Mg(NO3)2 166 for Cd and Pb, and Mg(NO3)2 for Cr determination. Ni was determined without the 167 addition of a chemical modifier. Experiments were performed to select the optimum 168 amounts of these modifiers. A series of measurements were performed to determine 169 the optimum quantities of both palladium and magnesium nitrate for As and Cu 170 determination. Therefore, a series of measurements were performed adding different 171 amounts of both modifiers. The amount of palladium varied from 0 to 6 μg and the 172 amount of Mg(NO3)2 varied between 0 and 5 μg. Taking into account the integrated 173 absorbance obtained and the peak profiles, amounts of 5 and 3 μg of palladium and 174 magnesium nitrate, respectively, were selected for As and Cu determination. Amounts 175 of NH4H2PO4 and Mg(NO3)2 were optimized for Cd and Pb determination. The amount 176 of NH4H2PO4 and Mg(NO3)2 required was studied in the range of 0–80 μg and 0–6 μg, 177 respectively. Taking into account the integrated absorbance obtained, 70 and 4 μg of 178 NH4H2PO4 and Mg(NO3)2, respectively, were selected as the optimum amounts for the 179 determination of both elements. The amount of Mg(NO3)2 varied from 0–40 μg for Cr 180 determination and 20 μg of this modifier was selected. 181 Influence of Acid Media on Slurry Preparation 182 After preparation of the slurry and to mobilize several elements into the solution, several 183 authors used an acid media. Thus, only a portion of the trace elements present 184 remained in the solid phase, which is nevertheless dispensed together with the 185 enriched solution into the atomizer during sampling (38,39). Several acids have been 186 reported in the literature for slurry preparation including HNO3 (8,9,40), a combination 187 of HNO3 plus H2O2 (41,42), HNO3, H2O2, plus HF (43), and HNO3 plus HClO4 (44). 188 HNO3 is widely used to prepare slurries of different types of samples such as seafood 189 (20,21), hair (22), plants (3), or sediments (23). Taking into account the results reported 190 in these reports, experiments were performed in our laboratory using HNO3 (containing 191 0.01% of Triton X-100) as a liquid media to prepare the slurry. The influence of the HNO3 192 concentrations was studied in the range of 0 to 1.0 M. The results obtained (Figure 1) 193 show that the integrated absorbance increased with the HNO3 concentration up to 0.2 194 M, and remained practically constant or slightly decreased at higher concentrations. 195 Therefore, in order to obtain the best analytical response, for the present study 0.2 M 196 HNO3 (containing 0.01% of Triton X-100) was selected as the liquid media to prepare 197 the slurry of terrestrial moss samples. 198 Calibration and Analyte Addition Graphs 199 To obtain calibration graphs of standard aqueous solutions containing concentrations 200 of 0–30 μg L-1 for As, 0–6 μg L-1 for Cd, 0–120 μg L-1 for Cu and Pb, 0–90 μg L-1 for Cr, 201 and 0–60 μg L-1 for Ni, and appropriate amounts of Triton X-100 solution, were added. 202 The standard addition method was used in the same range of concentrations using a 203 terrestrial moss sample. The slopes as well as the correlation coefficients (r) obtained 204 for the calibration and addition graphs are shown in Table III. Both, the calibration and 205 the addition graphs show good linearity, with correlation coefficients higher than 0.995 206 for all the elements studied. Moreover, both graphs exhibit similar slopes for all 207 elements. This means that aqueous calibration is a real possibility, and this type of 208 calibration was therefore used. 209 Sensitivity 210 Sensitivity was studied using three parameters: the limit of detection (LOD), the limit of 211 quantification (LOQ), and the characteristic mass (mo) (45). The LODs and LOQs 212 obtained are listed in Table IV. These limits were calculated based on 10 replicate 213 determinations of the blank. The instrumental LODs obtained varied from 0.031 μg L-1 214 for Cd to 0.94 μg L-1 for Ni. These limits are lower than those obtained by Serpil (43) for 215 the determination of these elements in lichen samples using ICP-OES (0.22, 2.3, 1.4, 216 2.3, and 7.4 μg L-1 for Cd, Cr, Cu, Ni and Pb, respectively). The LOD and LOQ values for 217 the moss sample, using 15 mg of sample and diluting to a final volume of 2 mL to obtain 218 the slurry, are also shown in Table V. These limits vary from 4.2 ng g-1 for Cd to 124.7 ng 219 g-1 for Ni. The characteristic mass (mo) is defined as the mass of the analyte (in 220 picograms) required to give a signal of 0.0044 for the integrated absorbance. The mo 221 values obtained were 31.0±6.2, 1.3±0.2, 8.5±1.6, 15.8±1.6, 24.4±0.4, and 33.1±3.6 pg 222 for As, Cd, Cu, Cr, Ni and Pb, respectively. The characteristic mass values are in 223 agreement with the manufacturer’s recommendation (37). 224 Precision and Accuracy 225 The within-run precision (relative standard deviation, RSD) of the method was obtained 226 for 10 replicate analyses of slurry of the internal reference material 227 (Pseudoscleropodium purum). The RSDs obtained were 4.5, 0.14, 3.7, 4.1, 2.2, and 228 3.7% for As, Cd, Cr, Cu, Ni and Pb, respectively. These values are lower than 10%, 229 indicating good precision. To study the accuracy of the method, three different slurries 230 of the certified reference material GBW07604 (for Cd, Cu, and Pb), M2 (As and Ni), M3 231 (Cr), and a blank were measured. The results listed in Table VI were compared with the 232 certified concentrations using a t-test (95% significance) (46). No statistical differences 233 between both concentrations were found; thus, it can be concluded that the method 234 suggested is accurate. The recovery was also calculated, and the mean recoveries 235 obtained were 101±0.9% for As (5, 10, 30 μg L-1 added), 103±6% for Cd (1, 2, 6 μg L-1 236 added), 96±2% for Cr (15, 30, 90 μg L-1 added), 106±1% for Cu (15, 30,6 μg L-1 added), 237 97±7% for Ni (10, 20, 60 μg L-1 added), and 106±2 % for Pb (30, 60, 120 μg L-1 added). 238 The recovery percentages were around 100% for all of the elements studied, therefore 239 the method shows good analytical recovery. 240 Application 241 The method was applied to the determination of As, Cd, Cu, Cr, Ni, and Pb in different 242 terrestrial moss samples collected around a ferrous smelter (Meg), a non-ferrous 243 smelter (Alc), and in a background area (OSE). Collection of the moss samples was 244 performed according to the methods developed for this moss species (48–50). The 245 collected samples were then prepared as described in the Sample Preparation section. 246 Two subsamples from each moss sample were prepared as slurry, and the analysis of 247 the slurry was performed in duplicate. The results in Table VII show that the As 248 concentrations varied from 205.8 (OSE) to 734.1 μg kg-1 and the arsenic concentrations 249 in the terrestrial moss samples collected around both smelters were two or three times 250 higher than the concentrations found in the background area. Cadmium concentrations 251 at OSE varied from 70.2 to 1142.2 μg kg-1 and the highest concentrations were found in 252 the samples collected around a ferrous smelter (Meg) (398–1142 μg kg-1). Chromium 253 concentrations (OSE) varied from 0.82–71.4 mg kg-1. Chromium concentrations in Meg 254 samples were from 4 to 87 times higher than the OSE sample (background area) and, in 255 general, the highest concentrations were found in the Meg samples. Copper and nickel 256 concentrations varied from 7.3 to 36.5 and from 2.1 to 36.5 mg kg-1, respectively. Lead 257 concentrations varied from 0.7 (OSE) to 165.5 (4Meg) mg kg-1. The highest lead 258 concentrations were found in Meg samples. 259 CONCLUSION 260 An environmentally friendly method for As, Cd, Cu, Cr, Ni, and Pb determination in 261 slurries of terrestrial moss samples by ETAAS was developed. To the best of our 262 knowledge, this is the first time that trace elements were determined in moss samples 263 using this methodology. This method decreases the number of sample pretreatment 264 steps (such as sample digestion) and avoids the use of corrosive reagents used in acid 265 digestion methods. This reduces sample preparation time (requires less than five 266 minutes) and reduces the possibility of sample contamination due to sample 267 manipulation and reagents used. The USS 800 titanium ultrasonic slurry sampler used 268 in this work is no longer commercially available. However, the slurry sample can be 269 prepared using an external ultrasonic probe and then introduced into the graphite tube 270 using the autosampler. The method was accurate and precise and presented high 271 sensitivity for the elements studied, with detection limits lower (LODs from 0.031 μg L- 272 1 for Cd to 0.94 μg L-1 for Ni) than those found in the literature (0.22, 2.3, 1.4, 2.3, 7.4 μg 273 L-1 for Cd, Cr, Cu, Ni and Pb, respectively) for the determination of these elements in 274 these types of samples (43). The method was applied to the determination of As, Cd, 275 Cu, Cr, Ni, and Pb in different terrestrial moss samples collected around a ferrous 276 smelter (Meg), non-ferrous smelter (Alc), and in a background area (OSE). Element 277 concentrations varied from 205.8 to 734.1 μg kg-1 for As; 70.2 to 1142.2 μg kg-1 for Cd; 278 0.82 to 71.4 mg kg-1 for Cr; 7.3 to 36.5 mg kg-1 for Cu; 2.1 to 36.5 mg kg-1 for Ni, and 0.7 279 to 165.5 mg kg-1 for Pb. These concentrations were similar to those reported in the 280 literature; moreover the analytical method allows determining concentrations in a wide 281 range of concentrations depending on the mass of the sample used for slurry 282 preparation. 283 References 284 1. R. Bargagli, Trace elements in terrestrial plants, Springer Verlag, Berlin, Germany, 285 324 pp (1998). 286 2. B. Smodiš, M.L. Pignata, M. Saiki, E. Cortés, N. Bangfa, B. Makert, B. Nyanko, J. 287 Arunachalam, J. Garty, M. Vutchokov, H. Th. Wolterbeek, E. Steinnes, M.C. Freitas, 288 A. Lucaciu, and M. Frontasyeva, J. of Atmospheric Chem. 49, 3 (2004). 289 3. R.A. Sánchez Moreno, M. J. Gismera, M.T. Sevilla and J.R. Procopio, Phytochem. 290 Analysis 21, 340 (2010) 291 4. A. Rühling and G. Tyler, Botaniska Notiser 121, 321 (1968) (published in English). 292 5. S. Giordano, P. Adamo, V. Spagnuolo, M. Tretiach, and R. Bargagli, Chemosphere 90, 293 292 (2013). 294 6. E. Steinnes, T. Berg and H. T. Uggerud, Science of the Total Environment 412, 351 295 (2011). 296 7. P. Adamo, P. Crisafulli, S. Giordano, V. Minganti, P. Modenesi, F. Monaci, E. Pittao, M. 297 Tretiach and R. Bargagli, Environmental Pollution 146, 392 (2007). 298 8. K. Bačeva, T. Stafilov, R. Šajn and C. Tănăselia, J. Environ. Science and Health Part 299 A, 47, 645 (2012). 300 9. I. Vučković, Z. Špirić, T. Stafilov, V. Kušan, J. Environn. Science and Health Part A, 48, 301 829 (2013). 302 10. J.R. Aboal, J.A. Fernández, T. Boquete, and A. Carballeira, Science of the Total 303 Environment 408, 6291 (2010). 304 11. B. Markert, O. Wappelhorst, V. Wecker, U. Herpin, U. Siewers, K. Friesek and G. 305 Breulmann, J. Raioanal. and Nuclear Chem. 240(2), 425 (1999). 306 12. P.C. Oniawa, Environmental Monitoring and Assessment 71, 13 (2001). 307 13. A. Ares, J.R. Aboal, A. Carballeira, S. Giordano, P. Adamo and J.A. Fernández, 308 Science of Total Environment 432, 143 (2012). 309 14. http://icpvegetation.ceh.ac.uk 310 15. H. Harmens, D. Norris and G. Mills, the participants of the moss survey, Heavy metal 311 and nitrogen in mosses: spatial patterns in 2010/2011 and long-term temporal 312 trends in Europe, Centre for Ecology and Hydrology, Bangor (UK), 63 pp (2013). 313 16. P. Viñas, N. Campillo, I. López-García and M. Hernández-Córdoba, Food Chem. 314 50(3), 317 (1994). 315 17. J.C. Latino, D.C. Sears, F. Portala and I.L. Shuttler, At. Spectrosc. 16(3), 121 (1995). 316 Table 7. 414 415